Arguments for creolisation in Irish English

Page created by Jim Pratt
 
CONTINUE READING
In: Raymond Hickey and StanisÓaw Puppel (eds) Language History and Linguistic
Modelling. A Festschrift for Jacek Fisiak on his 60th Birthday. Berlin: Mouton-de
Gruyter, 1997, 969-1038.

Arguments for creolisation in Irish English*
Raymond Hickey
University of Essen

1      Introduction
The area of creole studies has enjoyed up to the present undiminished interest among
linguists descriptive and theoretical alike and the results of this research has been of
considerable significance in the illumination of language variety over time and space and
has proved fruitful to considerations of language genesis.
         In the last decade or so historical stages of English, most notably Middle English,
have also been examined with a view to possible classification as a creole (Domingue
1977; Bailey and Maroldt 1979; Poussa 1982; Görlach 1986; Thomason and Kaufman
1988; Dalton-Puffer 1995; Danchev in press; on Romance, Schlieben-Lange 1979). The
claims of the pro-creolists, notably Bailey and Maroldt, have provoked strong reactions
from other linguists who have gone to considerable lengths to dismantle the creolisation
hypothesis for Middle English and northern varieties of Old English (cf. Thomason and
Kaufman 1988, for instance).
         Despite the adverse criticism which such attempts at attributing creole status to
historical varieties of English has met with I nonetheless deem it appropriate to consider
the question with regard to Irish English which has had quite a different history from
mainland English and where the synchronically ascertainable effects of sustained contact
with Irish justify at least an objective evaluation of possible creole status in the decisive
period of genesis for Irish English.
         In view of the intensity of reaction by opponents of creolisation and the negative
light in which the scholarship of the proponents is viewed I should stress at the outset that
I see no inherent advantage in classifying a stage of a language or variety as a creole.
Quite the opposite: both the terms ‘creole’ and ‘creolisation’ must be defined clearly as a
lack of terminological stringency leads to fuzziness which does a disservice to the field.
Given this standpoint let me begin with an attempt at definition and with an outline of the
main features which are a sine qua non for classification as a creole. Here I recognize
three main types of definition.

1.1    External definition

By this is meant that the criteria for definition refer to factors outside of the language,
ultimately to its sociolinguistic history. External definitions are favoured for instance by
such authorities as Holm (1988, 1994) when examining the independent varieties of
English in the Caribbean: ‘no particular set of syntactic features alone will identify a
language as a creole without reference to its sociolinguistic history’ (1994:372). Clearly
in a case such as this, where the political and demographic history of the region, with its
Raymond Hickey Arguments for creolisation in Irish English Page 2 of 55

large numbers of linguistically uprooted and displaced inhabitants, was highly conducive
to creole genesis, it is sufficient to refer to this history should there be dissenting voices
on the status of varieties of English in this region.
        However, definitions resting solely or even largely on external factors suffer
from one essential weakness, namely that it is ultimately a matter of opinion when one
regards the sociolinguistic environment as having been that for the origin of a creole.
This point is not usually the focus of attention of creolists as there is tacit agreement that
the non-standard varieties of certain key areas such as the Caribbean or West Africa are
accepted as creoles. Distinctions are then made on a vertical axis by lectal divisions or
the notion of cline (see below). Furthermore recognizably different varieties in a
geographical area may often be viewed as comprising a whole. Holm (1994) talks of
Creole (with a capital C) which would seem to imply some kind of unity if not
sociolinguistically at least internally as a structural complex.

1.2    Acquisitional definition

This type of definition sees a creole as a language which arises in a situation where a
generation of speakers develops its language from a drastically reduced and imperfectly
acquired form of a colonial lexifier language. Note that this definition is relational as its
point of reference is not a set of features of the language itself but a previous stage of the
lexifier and/or substrate language with which it is compared.
        The standard wisdom with types of definition acknowledges an historical
development from jargon (not stabilized or regular and with no fixed norms) to pidgin
and then to creole, the defining feature of the latter being the existence of speakers for
whom this is their native language. Of course to fulfill the needs of a first language a
pidgin has to be expanded considerably in grammar and lexicon in the process of
creolisation. The duration of the transition from jargon through to creole leads to
classifications for the interim stages such as ‘stable pidgin’, ‘expanded pidgin’, etc. The
possibility of a quick switch from jargon to creole (Bollée 1977) is captured by the term
‘abrupt creolisation’.

1.3    Structural definition

According to this definition a creole is a language which has undergone considerable
restructuring with respect to the lexifier language and probably with regard to the
substrate native language(s) as well (if such a language or languages provided input,
Versteegh 19??). This definition is also relational as it of necessity involves an
examination of the grammar of the lexifier language. However, if restructuring can be
narrowed down to an independent restricted set of features then it increases in value as
an internal definition.
         Restructuring can be taken to involve certain essential elements which, if missing,
diminish the status of a language as a creole. First and foremost among these is the
tendency towards analytical type. The simplification of the morphology which this
involves is frequently reached by utilizing common independent morphemes as
substitutes for bound morphemes in the grammar of the lexifier language. A clear
example is where a deictic element in pre-head position is used as the exponent of the
(inflectional) grammatical category and can be seen, for instance, with them as a plural
marker in English-based creoles. Pre-head position for grammatical elements is also
Raymond Hickey Arguments for creolisation in Irish English Page 3 of 55

seen in verbal phrases in the pre-verbal markers for tense, mood and aspect (more on this
presently).
        It should be remarked in this context that restructuring is to be distinguished
carefully from reanalysis or lack of analysis. The latter can arise from such phenomena
as a failure to recognize morpheme boundaries, e.g., to marrid; to fishin (Holm
1994:361) but is not to be confused with the active alteration of the grammar of the
incipient creole which represents restructuring.
        The remarks above refer to the lexifier language and not a possible substrate. Of
course there may be many structures which plausibly derive from a substrate in the
genesis of a creole and it is the relative weight accorded to the substrate which
constitutes the main difference between the universalist and substrate hypotheses in
creole studies (Muysken and Smith 1986, Alleyne 1986 and Mufwene 1986). In principle
the same processes may affect substrate and lexifier input to an incipient creole. One
respect in which restructuring is considered equally possible from above and below, so
to speak, is the promotion of aspect often with an attendant demotion of tense. This
feature is widely attested in those varieties/languages on which there is general
agreement as to their creole status, e.g., the various forms of imported English in the
Caribbean area.

1.3.1   Assessing status of change

For classification as a creole it is not enough just to have structural changes such as
movement to analytical type, particularly not if this can be accounted for by internal
developments which were active before the external situation which is supposed to have
lead to a change to creole. For instance the decline of inflections in the history of English
may well have been due to internal factors or earlier contact with Celtic speakers in
Britain (Hickey, 1995b); therefore this should this not be misinterpreted as an indication
of creolisation during the Norse and Norman French periods. Furthermore the lack of
pre-verbal modifiers and phonological simplification along with the continuing existence
of such syntactic features as the passive or clause subordination speak clearly against
creole status for Middle English. Those alterations in the grammar of English, such as the
changes in the system of personal pronouns, e.g., the borrowing in the north of forms in
th- from Scandinavian, can be accounted for internally and the rise of a form like she is
diametrically opposed to the tendency observed in many creoles where sex distinctions
are frequently not realised morphologically.
        Lexical borrowing is an indication of contact, direct or indirect, but can by no
means serve as a sign of creolisation. Hence the many day-to-day loans from
Scandinavian which entered English at the end of the Old English period, and the
occasional replacements to be found, point to close contact and mutual intelligibility but
not to an interruption in linguistic continuity which is a prerequisite for creole genesis.

1.4     Classificational difficulties

The consideration of the three main types of definition above illustrates the difficulty of
divising an absolute criterion for creole classification. In the opinion of the present
author the only two (related) characteristics which putative creoles can be said not to
share with other languages is that (i) they have arisen ex ovo in a relatively short time
within recorded history and (ii) that the linguistic input from which speakers created
these languages was not the native language of their parents.
Raymond Hickey Arguments for creolisation in Irish English Page 4 of 55

        There may well, however, be differences in the time scale: for example Jamaican
creole is a case with brief pidginisation, abrupt creolisation and for many speakers
gradual decreolisation and Tok Pisin is one with gradual expansion of an original pidgin
and later creolisation. The latter case would offer evidence for the gradualist view of
creolisation (Arends and Bruyn, 1994:111f.) which sees it as a process which takes
considerably longer than the generation or two assumed for abrupt creolisation. Creole
genesis does however imply a distinct interruption in the transmission of language in the
sense of a fully-fledged native language of a preceding generation. The process then
leads to nativisation, i.e., the creation of a creole, from a pidgin which, irrespective of its
degree of stabilisation, nonetheless shows alteration of its structures with the shift in
status, e.g., the switch from adverbials as indicators of aspect to a system of preverbal
markers.
        For the present paper there is no need to pursue the issue of how a variety arose
as a creole and the time span involved in this process any further. Rather it is more
germane to the matter at hand to distinguish carefully between classification as a creole
and the occurrence of structural elements typical of a creole. It is obvious when
considering a variety such as Irish English that it did not arise quickly (English has been
spoken in Ireland since the late Middle Ages, see below) so that scenarios of abrupt
creolisation are implausible for the Irish historical context.
        Although the outright classification of Irish English as a creole is rejected here
less absolute characterisations should be mentioned in this connection. First and foremost
is a notion which has been floated recently, Schneider’s ‘cline of creoleness’ (see
Schneider 1990 and Schneider this volume) a postulated scale along which creoles
would seem to differ in a non-discrete manner from a least standard to a more standard
form.1 The notion of ‘cline’ is intended to supersede the division of creoles into three
discrete entities basilect, mesolect and acrolect. For Irish English a cline of deviation
from southern British English can certainly be observed and probably had even greater
validity in previous centuries.

1.4.1   Other terms in the field

Obviously the term ‘pidgin’ refers to an indentifiable stage in the early development of a
new language, the later creole, so that when considering established languages, such as
English in its different varieties, from the point of view of similarities with
pidgins/creoles only the term ‘creolisation’, as a process with structural parallels to
creole genesis (see next section), would appear appropriate as there are always native
speakers of the varieties being considered. I will not examine the usefulness of related,
less established terms such as ‘creoloid’ or ‘semi-creole’ as such labels imply a steady
state of a variety at a level of intermediacy between a naturally transmitted variety of an
established language and a creole proper where it is impossible to reach agreement with
other scholars in the fields on just what are the defining features of such a half-way
house.

1.4.2   Creoles and creolisation

This is the essential distinction for the matter at hand; it is a distinction between an object
and the process which engendered it. A creole is a language with the characteristics
outlined above and discussed in detail below. Creolisation on the other hand is a process
whereby the universal features of language structure, unaffected by later counter-forces
which arise throughout the history of a language, come to the fore and forge the
Raymond Hickey Arguments for creolisation in Irish English Page 5 of 55

typological profile of a language. Traces of these features may still be discernible in a
variety in which case one can speak of a degree of creolisation in its history. For the
remainder of this paper I will thus be concerned with creolisation as a process operative
in the genesis of Irish English and with an outcome still visible in the present-day forms
of the variety. I am deliberately neglecting the question of the status of many structures in
particular their social values as this would necessitate another paper. For instance the
habitual aspect with do be is quite stigmatised but the immediate perfective with after is
universally accepted and is sociolinguistically unmarked.

1.5     Further matters of terminology

Outset and target language. These terms refer to Irish, a Q-Celtic language formerly the
majority language of Ireland, and English, the language switched to by the majority of the
Irish, essentially between the 17th and 19th centuries. The term ‘outset’ has two
advantages. It implies a temporal precedence of outset over target language which is
correct in the case of the shift from Irish to English and it also stresses that speakers start
from here acquisitionally without necessarily suggesting that the outset language was the
necessary source of any unexpected structures which appear in the target language which
is a definite implication embodied in the term ‘donor’ or ‘source’ language.

1.5.1   Terms for English spoken in Ireland

In the last two decades or so the term Hiberno-English2 (derived from Latin Hibernia
‘Ireland’ and English) has enjoyed a certain vogue. It would seem to have replaced the
older term Anglo-Irish although some authors, notably Patrick Henry and Loreto Todd,
have attempted to distinguish between the two (but arrive at opposite meanings for each
term). It is a label which is better avoided in linguistics as it is already used in literature
and politics and if taken literally would mean an English form of Irish (‘Anglo-’ as a
modifier to the head ‘Irish’). The term Irish English seems to the present author to be the
most neutral and least in need of explanation and has the additional advantage of being
parallel to other terms for varieties of English such as Welsh English, Canadian English,
etc. Further distinctions for sub-varieties can be introduced as required.3

2       The external history of Irish English
The involvement of England with Ireland is a long and complicated story. For the
purpose of the present article only a brief outline of English in Ireland is offered. The
first point to grasp is that there are two periods in the history of Irish English, a medieval
one and an early modern one.

2.1     The medieval period

The first period began in the late 12th century when Anglo-Norman adventurers landed in
the south-east of the country and established base there. There were English speakers in
their retinue, largely from the west of England, and these formed the core of the
communities of English settlers on the east coast from Waterford up to Dublin (Curtis
Raymond Hickey Arguments for creolisation in Irish English Page 6 of 55

1919, Cahill 1938). Later during the 13th century larger parts in the north and west were
subdued by the Normans after the initial invasion of 1169.
         Both the Anglo-Norman and English settlers of this initial period were
progressively assimilated by the native Irish so that by the end of the 16th century English
was reduced to only a small section of the population in the main cities and towns of the
east of the country and one or two rural pockets also in the east.
         There are very few linguistic documents from the early period. The main one is a
collection of poems called the Kildare Poems; besides these there are one or two other
small pieces and some municipal records chiefly from Waterford and Dublin. The
Kildare Poems are noted for their standardness. Given the fact that they were written in
an environment in which Irish or Norman French were the vital languages, one can
assume that the author or authors were deliberately writing what they regarded as a
supra-regional form of the English of the early 14th century (Hickey 1993).
         The turning point in the fortunes of English in Ireland is the establishment of the
House of Tudor in England (Moody and Martin 1994: 174-188); the Reformation which
they favoured led to a religious split between England and Ireland and the necessity to
subjugate the unruly Irish was deemed urgent to prevent Catholic rebellion spreading
from Ireland.

2.2     The early modern period

The re-establishment of English power got under way in the 17th century after the
decisive military defeat of the Irish in 1601 in Kinsale, Co.Cork. It attained a new quality
with the Cromwellian settlements of the 1650’s (Foster 1988:101-116). These were
undertaken to recompense mercenaries for services rendered during the political
struggles in England following the deposition of Charles I in 1649 and were much more
effective than the settlements in Munster (in the south) in the late 16th century.
        The plantations of the mid 17th century can be taken to mark the beginning of the
second, modern period of English in Ireland. All characteristics of contemporary Irish
English are taken to stem from the beginning of the early period, though there is evidence
in the phonology of popular Dublin English that elements survived from the late medieval
period. The difficulty with this contention lies in the paucity of documents from medieval
times so that for all questions of grammar no reference can be made to the first period.
        At this stage one should bear in mind that the development of the north and south
of the country had already begun to diverge. The key event for the province of Ulster was
the arrival of large numbers of settlers from Lowland Scotland as of the early 17th
century (after the political vacuum left by the exodus of Irish leaders in 1607, Moody and
Martin 1994:189-203). They were Protestant Presbyterians and spoke varieties of
Lowland Scots. The Scots tended to settle in the north-east (with a few other pockets)
and immigrants from northern England settled in the centre and south of the province.
Given the quantity of settlers and the fact that they were non-aristocratic settlers who
farmed the land and established towns as their bases their linguistic influence on the
local population was far greater than that of earlier settlers in the south. It is this Scottish
input which to this day is responsible for the very clear linguistic demarcation between
Ulster (in essence the state of Northern Ireland and a couple of adjacent counties such as
Donegal which are politically part of the Republic of Ireland) and the remainder of the
country to the south.
Raymond Hickey Arguments for creolisation in Irish English Page 7 of 55

2.2.1   The significance of the 17th century

The 17th century is of relevance to the theme of the present paper for a further reason.
Not only did speakers from the west of England settle in the south of Ireland but
substantial numbers of Irish moved voluntarily or under coercion to the New World. To
be precise, speakers of Scotch-Irish varieties of English emigrated from the north to
America in the 17th and 18th century and those from the south of the country were
transported to the Caribbean somewhat before. For instance Cromwell was responsible
for the shipment of native Irish to Barbados and Montserrat where the Irish worked as
indentured servants and gained the term ‘Black Irish’ or ‘Redlegs’ (Harlow 1926;
Sheppard 1977). There are many parallels, above all syntactic, between Irish English
and non-standard varieties of English in the Caribbean and some authors (Williams 1986
and 1988 for instance) would see these as evidence of Irish input in the formation of
Caribbean creoles. Others such as Rickford (1986) are more cautious in their assessment
of the role of Irish speakers of English. Again one must bear in mind that similarities
between creoles can as always be postulated as deriving from general structural
tendencies in creoles due to the nature of their genesis (more on which below).
         In the first half of the 19th century the rapidly increasing population of Ireland
(particularly in the south and west) led to a depletion of agricultural resources with
attendant famine culminating in the Great Famine of the late 1840’s (de Fréine 1966,
1977). During this event and for decades afterwards large-scale emigration took place, to
Britain (above all Merseyside) and to the United States. However the forms of southern
Irish English taken to America in the 19th century would appear not to have played any
role as input in the development of forms of American English.
         Allow me to return to the 17th century in the south of Ireland and the spread of
English from then onwards (O’Baoill 1990:150). A couple of general facts need to be
highlighted here.
         To begin with one should remark that when viewing the development of Irish
English it would be desirable to distinguish the English of native speakers of Irish from
that of the descendants of planters. Unfortunately there are no satisfactory means for
doing this. It is true that references exist to the English of the planters, for instance in the
caricatures of Swift (Bliss 1977), but leaving such minor sources aside, there is no body
of material which documents the English of those 17th century settlers after their arrival
in Ireland. The question of quantity is also of relevance here. The number of planters was
not that great so that their effect on the general development of Irish English should not be
overestimated. It is also unlikely that the language of the planters was affected by contact
with Irish speakers and that then the latter learned their English from later generations of
planters who spoke such a contact variety. This scenario, planter English - contact
variety - later acquisition by the Irish, does not pass muster as the number of planters was
too small for this to be likely. Furthermore for a contact variety to arise among planters,
they would have had to be in contact with English-speaking Irish for their English to be
affected by the broken English of this latter group. But how could rural Irish speakers of
English have predated the planters? What is more probable is that the Irish acquired
English slowly, altering it in the process and that with the general spread of this majority
variety the distinctiveness of planter English was lost so that there is no recognisable
derivative from this historical group today.
         Returning for a moment to the urban population, one can cite Kallen who in his
detailed survey of population distribution in the 17th century (1994:156-163) points out
that the concentration of English speakers in the towns would have meant that their
influence would have been greater than their mere numbers suggest. Exposure to English
Raymond Hickey Arguments for creolisation in Irish English Page 8 of 55

was also different in the towns and the countryside and this urban-rural split is still
evident (see Filppula 1991); there was also a degree of survival of English in the towns
of the east, above all Dublin, and this is responsible for the very obvious differences
between urban varieties of the east and the rest of the Republic of Ireland today. In
addition there was in the towns a large degree of bilingualism, indeed in Dublin in the
age of Swift (early 18th century) there were Anglophone and Gaelic literary circles
which existed side by side (Ó Háinle 1986).
         The manner in which the native Irish acquired English must also be taken into
account. There was obviously no formal instruction. Indeed education for the native
Catholics was forbidden under the Penal Laws which were not repealed until the end of
the 18th century. It must be assumed then that for the vast majority of rural Irish the
planters or other Irish were the source of English.
         This must have applied in the initial stages as then the rural natives would not
have had sufficient contact, if any at all, with those dwellers in the towns who would
have had English from the first period of settlement. This scenario in which a small
number of English planters conveyed the language to the native Irish would also explain
why the language of these planters had apparently been so strongly influenced by Irish:
the quantitative relationship was skewed in favour of the Irish so that the English could
not but be influenced by the numerically superior4 albeit socially inferior Irish (see
Thomason and Kaufman 1988:43+129). This kind of distribution also obtained in the late
12th and 13th centuries vis à vis the Anglo-Norman overlords who under pressure of
numbers (among other reasons) yielded to Irish and adopted it as their native language in
later generations (Hickey in press).
         The situation just sketched can be taken to have applied on a broad front for the
17th century. But the transmission of English to following generations of Irish was not
always directly from the settlers. Rather the Irish of the 18th and 19th centuries must have
learned English from Irish compatriots who were fortunate enough to have been exposed
to English, however imperfectly. Again one must stress that there was little if any formal
education. While it is true that so-called hedge schools existed whereby self-made
teachers gave instruction to Irish in a non-institutionalised manner (Dowling 1971) these
could not have serviced the entire country.
         Note that there is little or no distinction in present-day Ireland between those who
are descended from native Irish and those whose ancestors were planters in the 17th
century. This situation is markedly different from that in Northern Ireland where this
distinction is made, above all on religious grounds, that is it forms the basis for the
segregation of the Protestant and Catholic sections of the Northern Ireland community. In
the south there is a small Protestant section of the community which is definitely not
native Irish. The language of this section of the southern population does not, however,
differ significantly from that of the Catholic majority.5

2.3    Duration of shift

In summary allow me to stress the two main aspects of the development of Irish English
in the early modern period.

1)     A long switch-over period lasted from at least the mid 17th century to the second
       half of the 19th century and was characterised by extensive if poor bilingualism
       among the native Irish. Imperfect acquisition of English would have meant for a
       speaker of Irish that the latter would have had a continuing influence on the
Raymond Hickey Arguments for creolisation in Irish English Page 9 of 55

       former.6 In addition the acquisition of English was regarded as desirable for
       social advancement in Irish society.

2)     There was no displacement of population in Ireland; there was nothing like the
       mixture of (West African) backgrounds which characterised the forced
       immigrants in the Caribbean area in the early phase of the slave trade.
       Nonetheless there was a concentration of native population with the expulsion of
       the Irish to west of the Shannon (reflected in Cromwell’s dictum ‘to hell or to
       Connaught’) in the mid 17th century.

3)     The substrate language Irish was widely spoken and thus continually available.
       There was never a situation in which the Irish were deprived of a native
       language. Before the language shift was completed, English was acquired not for
       communication among the Irish but between these and English speakers.

The historical picture one is left with is that of a gradual dissemination of English from
east to west and from urban centres to rural districts over a period of at least two
centuries, from the Cromwellian era in the mid 17th century to the post-Famine period,
i.e., to the second half of the 19th century. Such a long period of bilingualism would have
furthered the transfer of structure from the outset language to the target one. The use of
speech habits and patterns from Irish on an individual level lasted long enough for these
to spread to the entire community of Irish speakers of English as general features of their
variety of the new language.

3      Access to English
The question of the access of the Irish to English is a thorny one. As indicated above in
the first period there was a small number of English immigrants who settled in the towns
of the east coast. The only genuine remnant of this period is the extreme south-east
corner, comprising the baronies of Forth and Bargy, and the area of Fingal just north of
Dublin. The evidence forthcoming on these traces of medieval English speak for a
radically altered variety. The glossaries of Vallencey (1788) and Poole (see Barnes
1867) are, judging by the aberrant orthography (Hickey 1988), quite unlike anything
spoken in Ireland today with the possible exception of popular Dublin speech.
         At the beginning of the early modern period access to this older type of English
would only have been available in the few urban centres. For the vast majority of Irish
living in the countryside there was no contact with indigenous Irish speakers of English.
Instead the exposure was gradual via those who had acquired English through contact
with urbanites and/or with the relatively small numbers of English people living in
Ireland or through self-education, the opportunities for which were minimal given the ban
on education for Catholics.
         When compared to historical situations in which pidgins arose the linguistic
plight of the Irish was probably better. For instance given the conditions of slavery in
which the deported Africans were kept in the Caribbean (Holm 1994:329) their access to
English was very restricted, basically consisting of the restructured English which their
immediate predecessors would have brought from West Africa, itself derived from
contact (direct or indirect) with English-speaking traders.7 The common assumption that
the children of slaves were exposed to ‘highly variable and possibly chaotic and
incomplete linguistic input’ (Holm 1994:330) did not apply to the Irish as they always
Raymond Hickey Arguments for creolisation in Irish English Page 10 of 55

had recourse to Irish. Later they only abandoned English after it had been acquired as a
native language. Thus the Bickertonian notion of an innate blueprint for language, his
biogram (see Bickerton 1984, 1988 and below) coming to the fore when there is
insufficient linguistic input, has never held at any stage in the development of Irish
English.

3.1    Role of superstrate varieties

When and for how long the Irish were exposed to English is one facet in the complex of
access to the language. The second and equally important one is what varieties of English
the Irish had as their input when acquiring the new language. Recent authors such as
Kallen and Harris have repeatedly pointed out that to ignore this question is to fail to
grasp a vital strand in the genesis of Irish English.
        Broadly speaking it is valid to maintain that western varieties of English
predominated among the kinds of English brought to the south from the beginning of the
17th century onwards. For the first period there is evidence of a south-western input (for
instance in the parallels between the dialect of Forth and Bargy and that of Dorset as
pointed out early on by the Dorset poet William Barnes who edited the main glossary in
this dialect). In the early modern period the extension of western input would seem to
have been further north, up as far as Lancashire.
        The western nature of early modern Irish English input is relevant when
considering the development or, at the very least, the reeinforcement of habitual
categories in Ireland. The west of England is an area which retained periphrastic do for
longer than did the north and east. Hence it is legitimate to assume that it was well
represented in the input to southern Irish English in the 17th century (see section 5.3.1.3.
Habitual below).
        There are other indications of western input. For instance the verbs have and be
usually have a single inflected form for the present, namely have and is.

(1)    a.      is for am, are, is                We’s up to our eyes in work.
       b.      have for have, has                She’ve a grand job at the glass.

Furthermore verbs tend to show a single form for preterite and past participle.

(2)    a.      When John come in, he sat down to the telly.
       b.      She done a lot of work in the house.
       c.      I seen him in town the other day.

This area of morphology shows a great deal of variety in the early modern period as
demonstrated quite clearly by Lass (1993) and for modern English by Cheshire (1993).
The matter is not simply that of the transition from strong to weak conjugational type but
what the manifestation and distribution of the strong verb forms are.
        Again many anomalies of Irish English can be attributed to the English input. For
instance Irish speakers frequently confuse verb pairs distinguished by direction such as
bring, take; rent, let; learn, teach. But this is a feature found with west English authors,
for example, with Shakespeare.

(3)    Caliban.        You taught me Language, and my profit on’t
                       Is, I know how to curse; the red-plague rid you
Raymond Hickey Arguments for creolisation in Irish English Page 11 of 55

                       For learning me your language!
                                                    (The Tempest, Act I, Scene 2)

However, not all verbal peculiarities of Irish English are immediately traceable to west
English input. One remarkable feature is seen in the epistemic use of negated must, noted
by Trudgill as distinctively Irish, (1986:140f.).

(4)    He musn’t be Irish as he was born in France.                Irish English
       He can’t be Irish as he was born in France.                 non-Irish English

In the pronominal area the distinction of first and second person plural can be safely
assumed to have been present in the input as may well have been the case with the dative
of disadvantage, as in She broke the glass on me.

3.2    The retentionist standpoint

Language is an internally structured system and its inherited established categories would
according to many authors have priority over contact as a source for idiosyncratic
features unless the evidence to the contrary is unambiguous. This standpoint has been
dubbed ‘retentionist’ (Filppula 1993:207-209) by authors dealing with Irish English and
its acceptance or dismissal forms a divide among the scholars of the field. The strongest
claim for the status of input varieties is Harris (1986:192f.) to the extent of dismissing
substratist views and backgrounding universalist explanations. Lass (1990a) is another
author who favours the retentionist view for the phonology of Irish English.8 A major
difficulty with the retentionist view as propounded by the authors mentioned is that it has
not been relativised9 by a sufficient consideration of both the external conditions under
which the Irish shifted slowly to English and of the parallels, phonological and syntactic,
which exist between Irish and English despite the obvious typological discrepancy
between the two languages. It is hoped to redress this imbalance in the relevant sections
to be found below.

3.3    Effects of supraregionalisation

The discussion above raises the question of what one is to regard as Irish English for an
investigation such as the present one. Clearly there are, or up to recently were, varieties
of Irish English which showed considerable influence from Irish, particularly in the
reported cases of literal translations of idioms from Irish. For instance Henry
(1977:33+36) and Adams (1983:11) provide many examples of sentences which are only
interpretable against the background of structure in Irish.

(5)    a.      The bate (beat) of him ishn’t in it.                (Irish English)
               Níl a bhualadh ann.
               [is-not his beat in-it]
               ‘He has no equal.’
       b.      The mother ishn’t too good to him.                  (Irish English)
               Níl an mháthair an-mhaith aige.
               [is-not the mother very-good at him]
               ‘His mother is not very well.’
Raymond Hickey Arguments for creolisation in Irish English Page 12 of 55

The scholars such as Harris, Kallen and Filppula who have (rightly) rejected the entirely
substratist hypothesis for Irish English frequently refer to such attestations to criticise a
slavishly contact view of the genesis of Irish English. Such instances nonetheless raise
the question of what is taken to constitute Irish English.
        For the present paper it is regarded as the supraregional vernacular variety of
English in the south of Ireland without undue concern for the issue of stigmatisation of
salient features of this variety. This is admittedly casting a wide and somewhat diffuse
net, but by taking the supraregional variety of the south one is detached from forms still
immediately affected by contact with Irish. Furthermore this supraregional variety has a
degree of stability, and one can assume that the features it shows (for instance its
aspectual categories) are central and not incidental to Irish English.

4      The contact case in Irish English
To open this section consider that between two or more languages there are many
possible situations of contact, four of which are listed below. These differ more in
degree than kind but are still discernibly different.

       Type                                      Effect
1)     Contact, but little if any                Only loan-words, cultural borrowings
       bilingualism
       (French in Middle English)

2)     Contact with approximation                Koineisation or dialect levelling,
       of one or both languages                  some structural permeation with similar
       to the other                              languages
       (late Old English and Norse)

3)     Contact with language shift               Grammatical interference, speech habits
       (Irish -> English)                        of outset transferred to target

4)     Contact but restricted input,             Pidginisation, grammatical restructuring,
       uncontrolled acquisition                  creolisation if there is no linguistic
       (Caribbean creoles)                       continuity of any kind

In this grouping Irish English occupies the third position but with leanings towards the
fourth this accounting for remarkable parallels between it and the languages subsumed
under 4).
        The distinction between interference and restructuring is essential here. The first
is a process of structural transfer between two languages and the second involves a
creative reorganisation of the grammar of a language on the basis of putative universals.
The latter will be considered below. Before this however the case for contact should be
dealt with.

4.1    What is unaltered on contact?

There would seem to be a general principle whereby the ‘deeper kernel’ of grammar in a
Raymond Hickey Arguments for creolisation in Irish English Page 13 of 55

language (Thomason and Kaufman 1988:5) is more resistant to change because it (above
all inflectional morphology) is so highly structured.10 For any highly structured
subsystem there is a standard wisdom that if it travels then only when it fits easily into
the recipient language. One is dealing here of course with closed classes; the lexicon on
the other hand is always open to incoming words and can equally donate words to
another language. However, it is a common error to conclude that a lack of numerous
loanwords critically weakens the case for any structural interference having occurred,
i.e., that there is no structural borrowing without lexical borrowing. This does not apply
to Irish English which has syntactic interference, through shift, but practically no lexical
loans from Irish.
         The resistance to structural influence is connected with the duration and extent of
contact. For instance long-term substratum interference can lead to a typological
reorientation of a language but within a time frame of several centuries at least. That is
definitely too large for the switch from Irish to English which was long enough for
considerable influence but not for a major typological realignment of English, especially
seeing that there was no linguistic discontinuity in Ireland.11

4.2    The notion of mixed language

A mixed language is taken to be one where different linguistic levels stem from different
languages. The most common configuration is for the morphology and syntax to derive
from one source and the lexicon from another. The latter, as an open class, can easily be
mounted or dismantled in the course of a language’s development and this phenomenon is
known in the history of creoles as re-lexification.
        The significant distinction between a mixed language and a creole is that the
grammar of the latter is created by speakers in an environment in which little grammar or
none is available as input and the former arises where for external reasons speakers
choose to combine linguistic levels from disparate sources. One of the languages may
have already existed independently as is the case with the well-known Media Lengua
which resulted from the re-lexification of Quechua with Spanish vocabulary (Appel and
Muysken, 1986:156f.). Ma’a (Mbugu, Goodman 1971) is another example of a mixed
language as it has Cushitic lexicon and Bantu morphology.
        This phenomenon is one which rests on typological similarity between inputs or
at least easy segmentability of the source grammar. For instance the adoption of Bantu
morphology was facilitated given the agglutinative and hence easily segmentable
character of Bantu grammar. Anglo-Romani (Hancock 1984) is an example from Britain
with Romani lexicon and English grammar.
        As a model for the genesis of Irish English the mixed language type must be
rejected seeing as how the Irish switched to English, altering it in the process, but did not
retain any of the grammar of Irish. It would not be sufficient to point to similarities in
syntax between outset and target in the Irish situation as a mixed language must show
actual elements from two languages and not as in Irish English the reflection of categories
from the outset language.

4.3    A code-switching model

Among the models for language mixing which have been the object of intensive research
in recent years is that of code-switching (see the representative selection of essays in
Raymond Hickey Arguments for creolisation in Irish English Page 14 of 55

Eastman 1992) and it would appear appropriate to consider this briefly within the
context of the present investigation. The best worked out model in this area is that of
Carol Myers-Scotton whose work has contributed much to understanding the
psycholinguistic mechanisms and structural constraints underlying code-switching and
borrowing, the latter being a phenomenon which she subsumes under the heading of
code-switching. For the languages involved in a code-switching situation Myers-Scotton
makes an essential distinction between matrix and embedded languages (1992:19-28).
The former is the language which provides the morphosyntactic frame and from which the
system units (morphemes of various kinds) stem and the latter donates lexical material
(content units) and possibly, under structural restrictions dictated by the matrix
language,12 morphological material as well.
         The following is a sketch of a code-switching situation which would have lead to
modern Irish English. It is a virtual scenario as there is no evidence for it. The
probability of it having corresponded to reality at any stage must be judged by the reader.
         Historically Irish would have been the matrix language from which most
morphology stemmed. Recall that the language which donates material to the matrix
language is termed the embedded language and this would originally have been English.
If one assumes this directionality then one is in essence postulating that Irish English is a
re-structured form of Irish which adopted English lexical material to fill the
morphosyntactic frames of Irish. There are one or two points of grammar where this view
is plausible, namely in the realisation of the perfective aspects. However the stance taken
by the present author is that the Irish switched to English by first acquiring it as a second
language and later by Irish being abandoned by following generations. The peculiarities
of this variety arose by the search for categorial equivalence among those Irish acquiring
English in an uncontrolled context.
         There are two major difficulties with a code-switching view of the genesis of
Irish English. The first is external. The motivation for code-switching is, apart from
obvious cases of lexical gaps, generally taken to be driven by the connotative and/or
indexical functions of the elements adopted into the matrix language. It is difficult to see
how many English lexical items would have fulfilled that requirement for the Irish as they
were not in contact with English society and would not have been au fait with non-Irish
social phenomena which they would then have glossed with English words via
code-switching in Irish.
         The second difficulty is that were Irish English restructured Irish then one would
expect its syntax to be much closer to Irish. Even in those instances where this is the
case, with the aspectual categories of Irish English, the word order is not that of Irish.13
And of course matters such as the post-specification of Irish have no equivalent in Irish
English nor is the heavily nominalised syntax reflected in the non-contact Irish varieties
of English. Finally there would have to have been total re-lexification to account for the
lack of Irish lexical items in Irish English (on this see section 5.5 below). However, the
historical attestations of Irish English from the late 16th century onwards offer no
evidence for any intermediary stages which would have necessarily existed in this
scenario.
         What one does observe in Modern Irish is code-switching within Irish (Stenson
1990, 1991) but not so much in the English spoken by native speakers of Irish (chiefly in
the surviving pockets of Irish in the west of the country). This code-switching is
characacterised by a high degree of morphological integration, for instance in the use of
the initial mutations and final palatalisation (markers of grammatical categories in Irish)
on code-switched lexical items. If this happens now then it would have happened three
centuries ago when Irish was in a much more vital position as the language of the
Raymond Hickey Arguments for creolisation in Irish English Page 15 of 55

majority of the population. This would then have meant that all this grammatical marking
would have had to disappear in a process similar to decreolisation. But again there are
no signs of this having happened in the history of Irish English.
        Among the literary parodies of Irish English which are available from the end of
the 16th century one has lexical sugaring, the sprinkling of words from Irish to lend a
flavour of genuineness to the language of these literary pieces. Again one must also bear
in mind that, apart from very few borrowings like tilly ‘small amount’ (from Irish
tuileadh) or twig ‘understand’ (from Irish tuigim) and the occasional use of Irish words
in English for special effect, such as plámás ‘flattery’, there is no measurable lexical
influence of Irish on English. There is nothing like the influence of French or
Scandinavian on English although the period of contact was considerably longer.

4.4     The typology of Irish

The points where transfer occurs is dependent on the structural match between the two
languages in question. In general it is true that Irish is typologically very different from
English. Irish not only shows differing surface features such as syntactic
post-specification (see section 6.1 below) but the entire organisation of its syntax is
unusual compared to English. The main difference is that Irish is a nominalising language.
Grammatical relationships and much of the semantics are expressed by means of nouns
and supporting prepositions (for introductory remarks see Hickey 1985a). Within the
scope of this paper it is not possible to go into this matter in detail. Suffice it to give one
example to convey the flavour of Irish syntax. Nouns with prepositions indicating
grammatical relationships are found where English has lexicalised verbs with an object.

(6)     a.      Rinne mé dearmad fúithi.
                [did I forget-noun under-her]
                ‘I forgot her.’
        b.      Níl teacht agat air.
                [is-not coming at-you on-it]
                ‘You cannot help it.’

There is a complex system of prepositional pronouns in Irish, some sixteen paradigms,
each of which consists of a directional preposition and an incorporated personal
pronoun. This system is not reflected in Irish English except in one single instance and
that is one where the prepositional pronoun in question does not show inflectional
variation but would seem to have taken on a lexical meaning, indicating existence14 and,
as an extension of this, possession.

(7)     a.      Tá morán daoine ann.
                [is lot people in-it]
                ‘There are a lot of people there.’
        b.      Níl a leithéid ann ar chor ar bith.
                [is-not his like in-it at all at all]
                ‘There is nothing to compare with him.’
        c.      Tá seans maith ann anois.
                [is chance good in-it now]
                ‘There is a good chance now.’
        d.      Tá seacht punt agam.
Raymond Hickey Arguments for creolisation in Irish English Page 16 of 55

               [is seven pounds at-me]
               ‘I have seven pounds.’

This use of ann ‘in-it’ to express existence has been transferred into Irish English and is
an almost stereotypical feature of Irish English.

(8)    a.      It’s bad weather that’s in it today.
               ‘The weather is bad today.’

Despite the major typological discrepancies between the two languages there are a few
remarkable parallels between Irish and English. One concerns the use of location as a
metaphor for action in progress.

(9)    a.      Fiche bliain ag fás.
               [twenty years at growing]
               Twenty years a-growing.

This fits neatly into English as can be seen from the translation of the above Irish
sentence. In fact there is a model for this in archaic forms of English as in I am on
writing for I am writing which with phonetic attrition of the preposition on gave the
prefixed schwa much as it did in phrases like asleep and alive from (much) earlier on
sl¤pe and on l»fe.
        A second parallel is the use of the same preposition on with a personal pronoun
to indicate relevance of an action.15

(10)   a.      They stole the car on me.
       b.      Ghoideadh an gluaisteán orm.
               [was-stolen the car on-me]

A third syntactic similarity is to be found in the positioning of the past participle. This is
placed after the object in Irish. There are good precedents for this in English and again
the practice of doing this may well have been present in the input varieties to Irish
English (as shown clearly by Harris 1983).

(11)   a.      I’ve the book read.
               Tá an leabhar léite agam.
               [is the book read at-me]
       b.      He’s the work done.
               Tá an obair déanta aige.
               [is the work done at-him]

A last parallel worthy of mention here is the lack of the for to-filter in Irish English
(Chomsky and Lasnik 1977, Henry 1995). This obviously has precedents in the history of
English and was probably not valid for the varieties of English brought to Ireland at the
beginning of the early modern period, let alone during the first period.

(12)   a.      Chuaigh sé go Baile Atha Cliath chun gluaisteán a cheannach.
               [went he to Dublin                  for car COMP buy]
       b.      He went to Dublin for to buy a car.
Raymond Hickey Arguments for creolisation in Irish English Page 17 of 55

The comments here concern the formal coincidence in both languages. The option of
placing the past participle either before or after the object has been functionalised in
Irish English and is clearly associated with perfective aspect, more on which below.

4.5     The search for categorial equivalence

If it is true that children in a situation of creole genesis have little or no grammar in their
linguistic input to fall back on, then it is equally true that in the language shift in Ireland
Irish formed the backdrop against which the incipient variety of English was moulded.
          If one reduces for a moment the situation of a community to that of the individual
then it would appear plausible that he/she would have expected and searched for
equivalents to the distinctions and structures present in the outset language Irish. This
phenomenon, which I postulate historically for Irish English and which is to be readily
observed in second-language acquistion, is what I term here ‘the search for categorial
equivalence’16 .
          Consider the following example. Irish has an immediate perfective which is
formed by the use of the prepositional phrase tar éis ‘after’ which is employed
temporally in this case.

(13)    a.      Tá siad tar éis an obair a dhéanamh.
                [is they after the work COMP do]

 The pivotal elements in this construction are listed below; the complementiser a is of no
semantic significance.

(14)    (i)     prepositional phrase tar éis
        (ii)    non-finite verb form déanamh
        (iii)   direct object obair.

Now it would appear that the Irish constructed an equivalent to this using English
syntactic means. Item (i) was translated literally as ‘after’, (ii) was rendered by the
non-finite V-ing form yielding structures as below.

(15)    They’re after doing the work.

Note that with a translation for (i) and a corresponding non-finite form for (ii) the task of
reaching a categorial equivalent would appear to have been fulfilled. Importantly the
word order Object + Verb was not carried over to English (*They’re after the work
doing). This illustrates a principle of economy: only change as much in L2 as is
necessary for other speakers in the community to recognize what L1-structure is intended.
       The structure dealt with above does not have any model in archaic or regional
English. With others however one can point to the existence of formal equivalents,
notably of the word order Non-finite Verb + Object. In this case there may well have
been an equivalent to the same Irish word order used to indicate a resultative perfective.

(16)    a.      Tá an leabhar léite aige.
        b.      He has the book read.

Although there were undoubtedly instances of the word order of the English sentence
Raymond Hickey Arguments for creolisation in Irish English Page 18 of 55

above, this does not mean that non-standard input varieties of English are in any way
responsible for the continuing existence of the word order in Irish English. This could
just as well have disappeared from Irish English as it has in forms of mainland English.
However the retention in Irish English and the use of the word order to express a
resultative perfective can be accounted for by the desire of Irish learners of English to
reach an equivalent to the category of resultative perfective which they had in their own
language.
         Among sociolinguists this phenomenon has been observed before. The additive
transfer of syntactic features to English can, for example, be captured neatly by Guy’s
notion of imposition (1990:49f.) whereby speakers in a language-shift situation impose
categorial equivalents to structures of the outset language onto the target language. In
traditional terms the process of calquing is similar as it involves as much translation into
the target language as is necessary for the structure to be recognised as an equivalent to
the substrate model by speakers during the process of language shift.
         Moving from the individual viewpoint to that of the community one can maintain
that if a majority of speakers hit on the same equivalents for outset structures and if there
is no external reason for rejecting these (such as stigmatisation) then such equivalents can
establish themselves as quasi-permanent features of the variety in question.

4.6    Neglect of distinctions

If the search for categorial equivalence has any validity then one would expect that the
reverse, the neglect of distinctions in the target but not in the outset language, would also
be found. This is indeed the case. Many authors have remarked on the relative
infrequency of the present perfect in Irish English (as early as Hume 1878, see Kallen
1990). As this category does not exist in Irish one can surmise that it was neglected by
learners of English in the period of language shift. Actions which began in the past and
continue into the present, or which are relevant to the present are expressed by the simple
present or past, whichever is appropriate. There may be some misinterpretation here as
with the following sentence which in more standard forms of English would refer to a
remote past.

(17)   a.      He never went to Dublin.
       b.      He’s never been to Dublin.

This phenomenon has been called the ‘extended present’ or ‘extended now’ (by Harris
and Kallen after McCoard 1978) and implies a greater use of the present than is normal
in other forms of English. But it can be viewed from two vantage points. Either it is an
extension of the present (the Irish perspective) or a neglect of the option of present
perfect (the English viewpoint). The Irish practice which has obviously been a
contributory factor can be illustrated by the following example.

(18)   a.      I know her since a long time now.
       b.      Tá aithne agam uirthi le tamall fada anois.
               [is knowledge at-me on-her with time long now]
       c.      He’s married for ten years.
       d.      Tá sé pósta le deich bliain.
               [is he married with ten years]
You can also read